src/gaia/utils/remove_duplicate_tests.py

#!/usr/bin/env python
"""
Script to remove duplicate test files.

This script identifies and removes duplicate test files based on predefined rules.
"""

import os
import logging
import shutil

# Configure logging
logging.basicConfig(
    level=logging.INFO,
    format="%(asctime)s - %(levelname)s - %(message)s"
)
logger = logging.getLogger("remove_duplicates")

# Define duplicate test files to remove
DUPLICATES_TO_REMOVE = [
    # Keep test_supabase_tables_comprehensive.py, remove others
    "tests/test_supabase_tables_final.py",
    "tests/test_supabase_tables_fix.py",
    "tests/test_supabase_tables_production.py",
    
    # Duplicate test files now consolidated
    "tests/test_gaia_agent.py",        # Consolidated into test_gaia_agent_consolidated.py
    "tests/test_fixed_agent.py",       # Consolidated into test_gaia_agent_consolidated.py
    "tests/test_duckduckgo_search.py", # Consolidated into test_duckduckgo_consolidated.py
    "tests/test_duckduckgo_integration.py", # Consolidated into test_duckduckgo_consolidated.py
    "tests/test_api_search.py",        # Consolidated into test_api_search_consolidated.py
    "tests/test_api_search_simple.py", # Consolidated into test_api_search_consolidated.py
    "tests/test_api_search_tool.py",   # Consolidated into test_api_search_consolidated.py
    "tests/test_api_search_tool_comprehensive.py", # Consolidated into test_api_search_consolidated.py
    "tests/test_api_search_standalone.py", # Consolidated into test_api_search_consolidated.py
    
    # Other potential duplicates
    "tests/minimal_supabase_test.py",  # Likely a duplicate or test version
]

def backup_file(file_path):
    """Create a backup of a file before removing it."""
    backup_dir = "backup_tests"
    if not os.path.exists(backup_dir):
        os.makedirs(backup_dir)
    
    # Get just the filename without the path
    filename = os.path.basename(file_path)
    backup_path = os.path.join(backup_dir, filename)
    
    # Copy the file to the backup directory
    shutil.copy2(file_path, backup_path)
    logger.info(f"Created backup of {file_path} at {backup_path}")

def remove_duplicate_files(file_paths, dry_run=False):
    """
    Remove duplicate test files.
    
    Args:
        file_paths: List of file paths to remove
        dry_run: If True, don't actually remove files
    """
    removed_count = 0
    
    for file_path in file_paths:
        if os.path.exists(file_path):
            logger.info(f"Removing duplicate test file: {file_path}")
            
            if not dry_run:
                # Create a backup before removing
                backup_file(file_path)
                
                # Remove the file
                os.remove(file_path)
                removed_count += 1
        else:
            logger.warning(f"File not found: {file_path}")
    
    return removed_count

def main():
    """Main function."""
    import argparse
    
    parser = argparse.ArgumentParser(description="Remove duplicate test files")
    parser.add_argument("--dry-run", action="store_true", help="Don't actually remove files")
    args = parser.parse_args()
    
    if args.dry_run:
        logger.info("Running in dry-run mode. No files will be removed.")
    
    # Remove duplicate files
    removed_count = remove_duplicate_files(DUPLICATES_TO_REMOVE, args.dry_run)
    
    # Print summary
    logger.info("\nCleanup Summary:")
    logger.info(f"- Duplicate test files removed: {removed_count}")
    
    if args.dry_run:
        logger.info("\nThis was a dry run. Run without --dry-run to apply changes.")

if __name__ == "__main__":
    # Force output to console even in dry-run mode
    handler = logging.StreamHandler()
    handler.setLevel(logging.INFO)
    formatter = logging.Formatter("%(asctime)s - %(levelname)s - %(message)s")
    handler.setFormatter(formatter)
    logger.addHandler(handler)
    
    # Add direct print statements for debugging
    print("Starting duplicate test removal script")
    print(f"Files to remove: {DUPLICATES_TO_REMOVE}")
    
    # Check if files exist
    for file_path in DUPLICATES_TO_REMOVE:
        if os.path.exists(file_path):
            print(f"File exists: {file_path}")
        else:
            print(f"File does not exist: {file_path}")
    
    main()
    
    print("Script completed")